Abstract
It has been argued that the willingness to pay for health care services in contingent valuation studies should be assessed
ex ante from an insurance perspective. It may however be easier to assess the willingness to pay among a group of patients in need of a specific treatment. This willingness to pay measure can be used to estimate the expected willingness to pay. This paper investigates the relationship between
ex ante and expected willingness to pay. It is shown that expected willingness to pay is a lower bound for
ex ante willingness to pay for a treatment that restores the individual to full health for an individual that is risk averse with respect to income. For a treatment that does not restore an individual to full health the expected willingness to pay is not necessarily a lower bound for the
ex ante willingness to pay if the marginal utility of income increases with better health.